Environ. Sci. Technol. 2004, 38, 4227-4232
Evaluation of Current Techniques for Isolation of Chars as Natural Adsorbents Y U A N C H U N , †,‡ G U A N G Y A O S H E N G , * ,† A N D CARY T. CHIOU§ Department of Crop, Soil, and Environmental Sciences, University of Arkansas, Fayetteville, Arkansas 72701, Department of Chemistry, Nanjing University, Nanjing 210093, People’s Republic of China, and U.S. Geological Survey, Box 25046, MS 408, Denver Federal Center, Denver, Colorado 80225
Chars in soils or sediments may potentially influence the soil/sediment sorption behavior. Current techniques for the isolation of black carbon including chars rely often on acid demineralization, base extraction, and chemical oxidation to remove salts and minerals, humic acid, and refractory kerogen, respectively. Little is known about the potential effects of these chemical processes on the char surface and adsorptive properties. This study examined the effects of acid demineralization, base extraction, and acidic Cr2O72- oxidation on the surface areas, surface acidity, and benzene adsorption characteristics of laboratoryproduced pinewood and wheat-residue chars, pure or mixed with soils, and a commercial activated carbon. Demineralization resulted in a small reduction in the char surface area, whereas base extraction showed no obvious effect. Neither demineralization nor base extraction caused an appreciable variation in benzene adsorption and presumably the char surface properties. By contrast, the Cr2O72- oxidation caused a >31% reduction in char surface area. The Boehm titration, supplemented by FTIR spectra, indicated that the surface acidity of oxidized chars increased by a factor between 2.3 and 12 compared to nonoxidized chars. Benzene adsorption with the oxidized chars was lower than that with the non-oxidized chars by a factor of >8.9; both the decrease in char surface area and the increase in char surface acidity contributed to the reduction in char adsorptive power. Although the Cr2O72oxidation effectively removes resistant kerogen, it is not well suited for the isolation of chars as contaminant adsorbents because of its destructive nature. Alternative nondestructive techniques that preserve the char surface properties and effectively remove kerogen must be sought.
Introduction The soil and sediment organic matter (SOM) has been known to act as the principal fraction for the soil/sediment sorption of nonpolar organic compounds (NOCs) from water (ref 1 * Corresponding author phone: (479)575-6752; fax: (479)575-3975; e-mail:
[email protected]. † University of Arkansas. ‡ Nanjing University. § U.S. Geological Survey. 10.1021/es034893h CCC: $27.50 Published on Web 06/29/2004
2004 American Chemical Society
and references therein). The NOC sorption to SOM occurs primarily by partition, characterized by a high degree of isotherm linearity and the lack of sorptive competition between coexisting solutes (contaminants) (2-4). Recent studies reveal, however, that the NOC sorption to natural solids containing a high-surface-area carbonaceous material, often referred to as black carbon (BC), may exhibit a nonlinear and competitive effect, which occurs usually at low relative concentrations (5-8). The BC results primarily from the field biomass burning that becomes embedded into soils and sediments. By this consideration, the enhanced isotherm nonlinearity of NOCs with the humin fraction of the Florida peat, as compared to the original peat and its humic acid fraction, was ascribed to the enriched BC in the humin fraction when fractionated with a density-fractionation procedure (9). Using the organic petrographic method, the BC particles have been visually identified and characterized (10-14). Theoretical models that combine adsorption by BC and partition into SOM have now been used to assess the contribution of BC in soils and sediments to NOC sorption (15-19). Confirmation of the adsorptive role of chars requires isolation, quantification, and characterization of the char properties. Geologists and limnologists have developed various techniques, including some destructive chemical treatment methods, for BC isolation and quantification on the basis that the BCs themselves are usually relatively inert to chemical treatment (e.g., refs 20 and 21). Common chemical treatments include acid demineralization, base extraction, thermal treatment, alternating base and organic solvent extraction, alkaline oxidation, and/or acidic oxidation (22-33). Demineralization using HCl and mixed HF-HCl solutions removes most carbonates and minerals with little formation of undesired precipitates (e.g., refs 22 and 28). Base extraction dissolves humic acid in high pH solutions, such as 0.1 M NaOH (e.g., ref 25). Thermal treatment burns off the organic matter insoluble in basic solution (e.g., kerogen) (15, 29-31), although complete removal of organic matter is not often achieved due to polymerization of the organic matter when heated in the presence of oxygen (27). Alternating base and organic solvent extraction removes much of kerogen but is too erratic to quantify BCs (25). Oxidation of kerogen with alkaline H2O2 has been used for the isolation of the aeolian BCs from marine sediments (22, 23, 32), but the treatment may not be generally effective due to the resistance of some kerogen to peroxide oxidation (33). Alternative to all these techniques, the acidic dichromate oxidation appears to be an effective means for removing kerogen on the recognition that kerogen is rapidly oxidized by dichromate while BCs are only slowly oxidized (26, 27). In 0.1 M Cr2O72- and 2 M H2SO4 at 50 °C, the observed halflife of kerogen removal is 6-180 h compared to 600-2000 h for BC removal (26). By carefully controlling reaction conditions and applying a correction for the loss of BCs, the dichromate oxidation technique appears adequate to quantify the BC content in soils and sediments. This technique now has been applied to BC quantification for various samples (34, 35). Whereas geologists and limnologists focus attention on BC quantification and age dating, environmental scientists are more interested in BC adsorptive properties. In a recent study, Song et al. (36) adapted a three-step procedure from the literature (i.e., demineralization, base extraction, and acidic dichromate oxidation) to obtain various particle fractions from soils, including BC, with the goal to characterize the adsorptive power of the indigenous BC for NOCs. VOL. 38, NO. 15, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4227
When attempting to evaluate the NOC uptake by an isolated BC, it is essential that the adsorptive power of the isolated BC after various chemical-treatment steps remains unchanged. Although extensive research shows that the above BC-isolation procedure accounts well for the BC mass, it is hypothesized that the procedure may cause the alteration of the BC surface properties, which would hinder our understanding of the behavior of BC as a natural adsorbent in soils and sediments. In this study, chars, one of the major kinds of BC, have been subjected to various chemical-treatment steps commonly adapted for BC quantification (26, 27, 30, 31, 33-36). They were those from the burning of wood and crop residue, the two primary contributors of environmental chars. After each step, the chars were characterized for their surface properties. Furthermore, some char-amended soils have also been treated in a similar manner to yield simulated, soil-derived chars for further characterization. The adsorption of a model NOC (benzene) from water by each char fraction was performed as additional sample characterization. The objectives of this study were to evaluate the potential impacts of various soil/solid chemical-treatment steps for BC isolation on the char surface and adsorptive properties and on the feasibility of such a procedure for substantiating the role of environmental chars in contaminant sorption and environmental fate.
Experimental Section Chars and Soils. Two laboratory-produced chars and a commercial activated carbon were used to represent char samples. Activated carbon, Darco G-60 (high-purity 100-mesh powder manufactured by American Norit Co.), was purchased from Aldrich Chemical Co. (Milwaukee, WI). Pinewood (Dacrydium L.) was collected from a private forest in Northwest Arkansas. Wheat residue (Triticum aestivum L.) was collected from the Arkansas Agricultural Research and Extension Center in Fayetteville, AR. Air-dried pinewood (100 g) and wheat residue (1000 g) were burned in the air on a stainless steel plate (1 m × 1 m) in an open field under natural conditions. The resulting charred materials, ∼25% and 6% of the weights of pinewood and wheat residue, respectively, were collected and pulverized to simulate chars produced by forest fires and burning of agricultural residues, respectively. Ash analysis indicated that charred pinewood contains only 0.4% ash, and therefore, it is assumed to consist of C as the major element, with minor elements being O and H. Charred wheat residue is found to contain, in addition to ∼16% of elemental C (i.e., char), high contents of Si and metal salts (37). To obtain the char fraction, the charred wheat residue (10 g) was treated in 200 mL of 1 M HCl solution four times and in 200 mL of HCl and HF (1 M:1 M) four times, followed by a thorough washing with distilled water four times to remove soluble salts and Si. Such purification has proven effective in the removal of Si and salts and in the enrichment of chars (37). The Stuttgart soil was collected at the Rice Research and Extension Center, Stuttgart, AR, with a composition of 17.1% sand, 60.4% silt, 22.5% clay, and 2.1% organic matter. The Houghton muck soil was collected at a muck farm in Lainsburg, MI, with an organic matter content of ∼90%. The soils, without records of crop residue burns, were assumed to contain minimal levels of chars. The soils were air-dried, ground, and sieved (1 mm). To obtain char-amended soils, pinewood char or wheat char (2.5 g) was added into the soils (47.5 g) in the exact (weight) content of 5% and thoroughly mixed before use for the experiments. Chemical Treatments. Chars and char-amended soils were subjected to demineralization, base extraction, and acidic dichromate oxidation using procedures identical to those described in Song et al. (36). Briefly, a char was demineralized by reactions with 6 M HCl at 60 °C for 20 h, 4228
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 15, 2004
with 22 M HF and 6 M HCl (2:1, v:v) at 60 °C for 20 h, and with 6 M HCl at 60 °C for 10 h, followed by a Soxhlet extraction for 72 h in a mixture of methanol, acetone, and benzene (2:3:5, v:v:v). The demineralized char was vacuum-dried at 60 °C to remove organic solvents; a portion (∼30% original weight) was stored in a vial. The char was further subjected to base extraction in N2 gas-purged 0.1 M NaOH for 12 h; this process was repeated five times for a total of 60 h. The char was freeze-dried, and a second portion (∼30% original weight) was stored. Finally, the remaining char was oxidized in 0.1 M K2Cr2O7 plus 2 M H2SO4 at 55 °C for 60 h; the solution of K2Cr2O7 and H2SO4 was replenished twice during oxidation. Following a thorough washing with distilled water, the char was dried at 60 °C. The char-amended soils were subjected to the identical chemical treatment, except that the solution of K2Cr2O7 and H2SO4 was replenished four to six times during oxidation; final residues were obtained and weighed. The entire procedure resulted in a total of 16 chars, i.e., three starting chars, three acid-demineralized chars, three baseextracted chars, three oxidized chars, and four soil-removed oxidized chars. These chars are referred to as AC, PC, and WC for the starting activated carbon, pinewood char, and wheat residue char, respectively, as AC-A, PC-A, and WC-A after acid demineralization, as AC-AB, PC-AB, and WC-AB after base extraction, and as AC-ABO, PC-ABO, and WCABO after Cr2O72- oxidation. The soil-removed oxidized chars are referred to as PC-S-ABO and WC-S-ABO as well as PCM-ABO and WC-M-ABO, respectively, for PC- and WCamended Stuttgart and Muck soils. Surface Areas. The surface areas were measured by N2 adsorption at liquid nitrogen temperature using a Gemini2360 Micromeritics surface area analyzer. Chars (∼0.2 g) were outgassed overnight (∼15 h) at 105 °C under a helium gas flow at 20 mL/min before N2 adsorption. The molecular surface area of 16.2 Å2 for N2 and the BET (BrunauerEmmett-Teller) equation were used to calculate the surface areas of the chars. Surface Acidity. The oxygenated acidic surface groups of chars were determined using Boehm’s titration method (38, 39). Prior to measurement, the chars were equilibrated with dilute HCl solution at pH 2 for 3 d, followed by a thorough washing with deionized water until they were free of Cl- as detected by AgNO3. A dried char (0.2-0.5 g) was mixed with 25 mL of 0.05 N NaHCO3, while a replicate char was mixed with 25 mL of 0.05 N NaOH, and both chars were shaken for 24 h. Excess base in 5 mL of the filtrate was titrated with 0.01 N HCl. Surface acidity was calculated based on the assumption that NaHCO3 neutralizes carboxyl groups only and NaOH neutralizes all acidic groups including carboxyl, lactonic, and phenolic groups. Fourier Transform Infrared Spectroscopy (FTIR). AC and AC-ABO and WC and WC-ABO in KBr wafers (2.0%) were prepared for FTIR measurements. The diffuse-reflectance IR spectra were obtained using a Nicolet Impact 410 FTIR spectrometer (Nicolet Instrument Corp.) with 4 cm-1 resolution and 64 scans between wavenumbers of 700 and 2000 cm-1. The instrument was run with the AC sample as the background, and the spectra for all other chars were obtained by subtracting the background. NOC Adsorption. Benzene, a common neutral contaminant and widely used in many studies, was selected as a model NOC. Benzene adsorption by chars was measured by the batch equilibration method, as described in many earlier studies (e.g., 8). A char (∼0.02 g) was mixed with 4 mL of deionized water in 5-mL glass vials for 24 h to allow the surface of the char to hydrate. Various volumes of neat benzene (0.7-5.6 µL) were delivered into the vials using a Hamilton microsyringe. The vials were immediately closed with Teflon-lined screw caps and rotated (40 rpm) at room temperature (∼25 °C) for 48 h. Previous study showed that
benzene adsorption by BC reached apparent equilibrium within 24 h (37). After adsorption equilibrium, solid and aqueous phases were separated by centrifugation at 6000 rpm (RCF ) 5210g) for 20-30 min. A volume of 2 mL of supernatant was extracted with 4 mL of hexane in a glass tube by agitating for 60 min on a reciprocating shaker. A portion of the hexane phase containing the extracted benzene was then analyzed using gas chromatography. Blank samples not containing chars were also prepared and analyzed using the same procedure. The measured recovery was ∼97%. Concentrations were adjusted for the recovery. The amount of benzene adsorbed was calculated from the difference between the amount added and that remaining in the final solution. All measurements were in duplicate, with the difference being generally WC, consistent with the decreasing order of surface areas and increasing order of surface acidity of the chars. The chars subjected to demineralization (AC-A, PC-A, and WC-A) and base extraction (AC-AB, PC-AB, and WC-AB) show no appreciable differences in benzene adsorption, as these chars have surface areas and surface acidity similar to those of the starting chars. This finding suggests the suitability of acid demineralization and
FIGURE 2. Isotherms of benzene adsorption by starting chars (AC, PC, and WC representing activated carbon, pinewood char, and wheat char, respectively) and those subjected to chemical treatments in the order of acid demineralization (AC-A, PC-A, and WC-A), base extraction (AC-AB, PC-AB, and WC-AB), and Cr2O72oxidation (AC-ABO, PC-ABO, PC-S-ABO, PC-M-ABO, WC-ABO, WCS-ABO, and WC-M-ABO). base extraction steps for char isolation. In contrast, the chars subjected to the Cr2O72- oxidation show a much reduced benzene uptake as a result of the sharp decrease in surface area and the dramatic increase in surface acidity of these chars. The chars isolated from the char-amended soils show a similar benzene adsorption. Therefore, the Cr2O72- oxidation is clearly not appropriate for the preservation of the char adsorptive power. We are unable to compare the present benzene adsorption with the NOC adsorption by BCs isolated by Song et al. (36) from soils and sediments since such data have not been reported by these scientists. Table 1 also presents the benzene adsorption, determined from the isotherms, at the equilibrium concentration of 180 mg/L (relative concentration of 0.1) for all chars. The adsorption data with the starting chars and the oxidized chars are chosen for comparison. The adsorption data with the chars isolated from the char-amended soils may be complicated by the possible presence of residual SOM on the surfaces of the chars. Careful examination reveals that benzene adsorption by AC-ABO compared to that by AC is reduced by a factor of 8.89 while the reduction in the surface area of the corresponding samples is by a factor of only 1.84. Similar results are also found with PC and WC series, where the respective factors are 33.8 and 13.1 for the former and 16 and 4.25 for the latter. A larger reduction in benzene adsorption than in surface area with the Cr2O72- oxidation of chars manifests the influence of surface acidity on the
overall solute adsorption. This effect may also be seen by comparing surface-area-normalized benzene adsorption values, QSA (mg/m2), calculated from the benzene adsorption at 180 mg/L in Table 1 divided by the surface area of the corresponding char. Normalization to char surface area of benzene adsorption eliminates the surface area effect. The QSA values are 0.267 and 0.0552 mg/m2 for AC and AC-ABO, 0.232 and 0.0897 mg/m2 for PC and PC-ABO, and 0.128 and 0.0342 mg/m2 for WC and WC-ABO, respectively. They are 3-5-fold lower for the oxidized chars than for the corresponding starting chars or the demineralized and baseextracted chars (data not shown). Assuming that the lower QSA is due solely to increased surface acidity, the contribution of the acidity effect to the total benzene adsorptive reduction is 79.3% for AC-ABO, 61.3% for PC-ABO, and 73.3% for WCABO. Clearly, the increase in surface acidity of chars, relative to the reduction in surface area, both caused by the Cr2O72oxidation, played a predominant role in the benzene adsorptive reduction. In conclusion, chars in soils and sediments may significantly contribute to NOC sorption. Understanding their role relies on reliable techniques of isolation. Although acid demineralization may reduce the char surface area to a certain degree, it does not result in an increased surface acidity and a significantly decreased NOC adsorption. Base extraction does not affect the char surface area, surface acidity, and NOC adsorption. Demineralization and base extraction appear to be proper for removing salts, minerals, and humic acid. Acidic dichromate oxidation causes a large reduction in surface areas and an increase in surface acidity of chars. As a consequence, the NOC adsorption by chars is dramatically reduced. It is clear that chars isolated from environmental matrixes via destructive oxidation are not appropriate for adsorptive characterization. As current isolation techniques involve oxidation processes, new nondestructive techniques that do not cause substantial alteration of char surface while effectively removing kerogen are much in order.
Acknowledgments This research was supported by USDA-NRICGP Grant 200235107-12350 and the University of Arkansas Division of Agriculture. The use of trade and product names is for identification purposes only and does not constitute endorsement by the U.S. Government.
Literature Cited (1) Chiou, C. T. Partition and Adsorption of Organic Contaminants in Environmental Systems; John Wiley & Sons: Hoboken, NJ, 2002. (2) Chiou, C. T.; Peters, L. J.; Freed, V. H. Science 1979, 206, 831832. (3) Chiou, C. T.; Shoup, T. D.; Porter, P. E. Org. Geochem. 1985, 8, 9-14. (4) Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Water Res. 1979, 13, 241-248. (5) Weber, W. J.; Huang, W. Environ. Sci. Technol. 1996, 30, 881888. (6) Weber, W. J.; McGinley, P. M.; Katz, L. E. Environ. Sci. Technol. 1992, 26, 1955-1962. (7) Xing, B.; Pignatello, J. J. Environ. Sci. Technol. 1997, 31, 792799. (8) Xing, B.; Pignatello, J. J.; Gigliotti, B. Environ. Sci. Technol. 1996, 30, 2432-2440. (9) Chiou, C. T.; Kile, D. E.; Rutherford, D. W.; Sheng, G.; Boyd, S. A. Environ. Sci. Technol. 2000, 34, 1254-1258. (10) Kleineidam, S.; Rugner, H.; Bertaind, L.; Grathwohl, P. Environ. Sci. Technol. 1999, 35, 1637-1644. (11) Karapanagioti, H. K.; Kleineidam, S.; Ligouis, B.; Sabatini, D. A.; Grathwohl, P. Environ. Sci. Technol. 2000, 34, 406-414. (12) Karapanagioti, H. K.; Childs, J.; Sabatini, D. A. Environ. Sci. Technol. 2001, 35, 4684-4690. (13) Ghosh, U.; Seb Gillete, J.; Luthy, R. G.; Zare, R. N. Environ. Sci. Technol. 2000, 34, 1729-1736. VOL. 38, NO. 15, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4231
(14) Karapanagioti, H. K.; Sabatini, D. A. Environ. Sci. Technol. 2000, 34, 2453-2460. (15) Gustafsson, O ¨ .; Haghseta, F.; Chan, C.; MacFarlane, J.; Gschwend, P. M. Environ. Sci. Technol. 1997, 31, 203-209. (16) Xia, G.; Ball, W. P. Environ. Sci. Technol. 1999, 33, 262-269. (17) Accardi-Dey, A.; Gschwend, P. M. Environ. Sci. Technol. 2002, 36, 21-29. (18) Accardi-Dey, A.; Gschwend, P. M. Environ. Sci. Technol. 2003, 37, 99-106. (19) Allen-King, R. M.; Grathwohl, P.; Ball, W. P. Adv. Water Resour. 2002, 25, 985-1016. (20) Swain, A. M. Quat. Res. 1973, 3, 383-396. (21) Goldberg, E. D. Black Carbon in the Environment; Wiley: New York, 1985. (22) Herring, J. R. In The Carbon Cycle and Atmospheric CO2: Natural Variations Archean to Present; Sundquist, E. T., Broecher, W. S., Eds.; American Geophysics Union: Washington, D.C., 1985; pp 419-442. (23) Griffin, J. J.; Goldberg, E. D. Limnol. Oceanogr. 1975, 20, 456463. (24) Hayatsu, R.; Matsuoka, S.; Scott, R. G.; Studier, M. H.; Anders, E. Geochim. Cosmochim. Acta 1977, 41, 1325-1339. (25) Wolbach, W. S.; Lewis, R. S.; Anders, E. Science 1985, 230, 167170. (26) Wolbach, W. S.; Anders, E. Geochim. Cosmochim. Acta 1989, 53, 1637-1647. (27) Lim, B.; Cachier, H. Chem. Geol. 1996, 131, 143-154. (28) Smith, D. W.; Griffin, J. J.; Goldberg, E. D. Anal. Chem. 1975, 47, 233-238. (29) Gelinas, Y.; Prentice, K. M.; Baldock, J. A.; Hedges, J. I. Environ. Sci. Technol. 2001, 35, 3519-3525. (30) Cachier, H.; Bre´mond, M. P.; Buat-Me´nard, P. Tellus 1989, 41B, 379-390.
4232
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 15, 2004
(31) Ducret, J.; Cachier, H. J. Atmos. Chem. 1992, 15, 55-67. (32) Smith, D. W.; Griffin, J. J.; Goldberg, E. D. Nature (London) 1973, 241, 268-270. (33) Wolbach, W. S.; Anders, E.; Nazarov, M. A. Geochim. Cosmochim. Acta 1990, 54, 1133-1146. (34) Masiello, C. A.; Druffel, E. R. M. Science 1998, 280, 1911-1913. (35) Masiello, C. A.; Druffel, E. R. M.; Currie, L. A. Geochim. Cosmochim. Acta 2002, 62, 465-472. (36) Song, J.; Peng, P.; Huang, W. Environ. Sci. Technol. 2002, 36, 3960-3967. (37) Yang, Y.; Sheng, G. Environ. Sci. Technol. 2003, 37, 3635-3639. (38) Boehm, H.-P. Adv. Catal. 1966, 16, 179-274. (39) Boehm, H.-P.; Diehl, E.; Heck, W.; Sappok, R. Angew. Chem., Int. Ed. 1964, 3, 669-677. (40) Bansal, R. C.; Donnet, J. B.; Stoeckli, F. Active Carbon; Marcel Dekker: New York, 1988. (41) Leon y Leon, C. A.; Radovic, L. R. In Chemistry and Physics of Carbon; Thrower, P. A., Ed.; Mercel Dekker: New York, 1994; Vol. 24, pp 213-310. (42) Guo, Y.; Bustin, R. M. Int. J. Coal Geol. 1998, 37, 29-53. (43) Foley, N. J.; Thomas, K. M.; Forshaw, P. I.; Stanton, D.; Norman, P. R. Langmuir 1997, 13, 2083-2089. (44) Mu ¨ ller, E. A.; Rull, L. F.; Vega, L. F.; Gubbins, K. E. J. Phys. Chem. 1996, 100, 1189-1196.
Received for review August 12, 2003. Revised manuscript received May 25, 2004. Accepted May 26, 2004. ES034893H
